A recent warning from researchers over concerns that AI could be harmful and destructive has prompted a reassessment of the technology's future — and added urgency to a growing debate over how, or whether, to regulate it.
President Donald Trump has pushed back on those concerns, arguing the federal government already has "tremendous criminal and regulatory power" over AI companies and that additional restrictions could jeopardize U.S. competitiveness, particularly against China.
"The only control or 'guardrails' that AI needs is a STRONG AND SMART (High IQ!) PRESIDENT, and the U.S.A. has that, in spades!" Trump wrote in a social media post Monday.
But leaders within the AI industry itself have signaled alarm. Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman have both recently called for additional safeguards, including government involvement and independent oversight.
"No amount of American competitive pressure should justify recklessness, or let capabilities get ahead of alignment and monitoring," Altman said in a post on X Sunday night.
Anthony Aguirre, president and CEO of the Future of Life Institute — a nonprofit focused on steering transformative technologies away from large-scale risks — recently convened what he called a "pro-human assembly" in Washington, bringing together figures from across the political spectrum, including Bernie Sanders and Steve Bannon.
RELATED STORY | Bernie Sanders and Steve Bannon find rare common ground: AI needs the brakes pulled
"All these people have something in common. They're people. And so there are these common human interests that I think really, you know, people start to see that they're threatened really can unite those people behind them," Aguirre said.
The assembly produced what organizers call the Pro-Human Declaration, a document now signed by more than 1 million people. Its principles range from requiring disclosure when someone is interacting with an AI system to ensuring AI remains under meaningful human control.
Aguirre said strong regulation requiring developers to demonstrate control over their systems before deployment could prevent many of the risks posed by advanced AI.
"If we had strong regulation that says before you build an AI system, you got to demonstrate to someone, you know, someone who can say no, that you've got the systems in place to be able to control that thing and keep oversight over it and make sure it doesn't go rogue. That's going to stop a lot of the problems that we're having with these advanced, powerful AI systems," Aguirre said.
On the question of whether regulation would cede ground to China, Aguirre argued the framing is flawed.
"Nobody benefits from creating an AI system that's out of control. We don't benefit. The Chinese don't benefit either," Aguirre said.
He also addressed skepticism over why major AI companies are calling for guardrails, noting that a recent incident — in which a collective of AI agents reportedly hacked out of one company and into another without detection for months — may have changed the calculus for the industry.
"I think they felt like they had this under control and then they saw a collective of AI rogue agents hack out of their own company and into another company that they didn't know was going on for months," Aguirre said.
RELATED STORY | AI regulation stalls in Congress as House votes on data center costs
Aguirre said the goal is not to stop AI development, but to redirect it.
"We just need to not build the scary rogue things and build the sort of tools that everyday people actually want to make their lives better and empower them to do more things," Aguirre said.
